PURPOSE: to determine the main factors associated with vaginal delivery in high-risk pregnant women submitted to labor induction with vaginal misoprostol (50 µg). METHODS: this is a secondary analysis of an open nonrandomized clinical trial that included 61 high-risk pregnant women admitted at the "Maternidade-Escola Assis Chateaubriand", Fortaleza (Ceará). All women had singleton pregnancies with alive fetuses, gestational age >37 weeks and Bishop scores <7. Misoprostol was vaginally administered at doses of 50 µg every 6 h for a maximum of four doses. Univariate and multiple logistic regression analyses were performed to determine association between vaginal delivery (dependent variable) and independent variables (predictive), and receiver operating characteristic (ROC) curves were constructed for parity and Bishop scores. RESULTS: parity (one or more previous deliveries), Bishop scores >4 and interval induction to delivery <6 h were significantly associated with vaginal delivery, while tachysystole reduced the probability of vaginal delivery. A multivariate stepwise logistic regression was then performed to evaluate each of these as independent predictors. Parity (OR = 5.41, 95% CI = 4.18-6.64) and Bishop score >4 (OR = 3.30, 95% CI = 2.15-4.45) were significant independent predictors for vaginal delivery. In the ROC curve for parity and Bishop score, sensitivity of 63.2% and positive predictive value of 100% were found. The area under the ROC curve was 86.8%, significantly higher than 50% (p=0.023). CONCLUSIONS: the most important predictive factors for vaginal delivery after induction with misoprostol were parity and Bishop score. These characteristics should be considered when choosing schemes and doses of misoprostol for cervical ripening and labor induction.
Labor induction; Prostaglandins; Labor
